Add turbomind metrics

[lvhan028]

Thanks for your contribution and we appreciate it a lot. The following instructions would make your pull request more healthy and more easily receiving feedbacks. If you do not understand some items, don't worry, just make the pull request and seek help from maintainers.

Motivation

Implemented metric collection for scheduler and inference in TurboMind engine ( Resolve #3801)

• Metric collection now covers both scheduling and inference processes
• Modified files: csrc/*, bind.cpp, and turbomind.py

Optimized metric polling mechanism

• Switched from push to pull model for scheduler metrics, as scheduler metrics change less frequently than per-token metrics

Corrected metrics statistics

• Tokens per second (TPS)
• Inference time latency (ITL)

Log condition

• Metrics are logged to console only when requests are received.

[CUHKSZzxy]

LGTM